Cerebral Hemorrhage and Amyloid-β

A recent article by M. Pfeifer and colleagues (“Cerebral hemorrhage after passive anti-Aβ immunotherapy,” Brevia, 15 Nov., p. [1379][1]) reports that the removal of deposited amyloid-β from the vasculature leads to increased cerebral microhemorrhages. These results suggest that it is not the deposition of amyloid-β, but rather the lack of deposited amyloid-β that could lead to cerebral hemorrhage and stroke. Taken another way, the deposition of amyloid-β in the first place may be a mechanism to prevent cerebral hemorrhage. Unlike the transgenic mouse brain and vasculature, where amyloid-β deposition is simply a result of the overexpression of amyloid-β protein precursor, amyloid-β deposition during adult life is normally associated with injury ([1][2]). Mounting evidence indicates that this deposition of amyloid-β may be a neuroprotective response to injury ([2][3], [3][4]). Given the extensive vasculature of the brain and that peripheral mechanisms of preventing hemorrhage (i.e., platelet closure) would prevent nourishment of neurons, other mechanisms have likely evolved to rapidly seal vascular ruptures without activating the coagulation cascade. We have proposed that amyloid-β is one such sealant, given its physiochemical properties that allow it to aggregate under inflammatory conditions ([3][4], [4][5]) and form an intracranial “scab,” thereby maintaining structural integrity of the blood-brain barrier. Importantly, such a notion explains the acute phase generation and rapid cortical deposition of amyloid-β in stroke and after head trauma ([1][2]) and its resolution after recovery ([5][6]), important physiological responses that would limit the loss of terminally differentiated neurons. Removal of such a seal would lead to hemorrhage and an inflammatory immune response ([4][5]). Indeed, the results of Pfeifer et al. suggest that the root cause of the encephalitis and meningitis suffered by individuals in the Elan immunotherapy trial is directly associated with the removal of amyloid-β from the vasculature. These results demonstrate yet again the futility of removing a protein, amyloid-β, which has ubiquitous tissue expression, without first understanding its function(s). 1. [↵][7]1. G. W. Roberts 2. et al. , J. Neurol. Neurosurg. Psychiatry. 57, 419 (1994). 2. [↵][8]1. K. Iqbal, 2. S.S. Sisodia, 3. B. Winblad 1. C. S. Atwood 2. et al. , in Alzheimer's Disease: Advances in Etiology, Pathogenesis and Therapeutics, K. Iqbal, S.S. Sisodia, B. Winblad, Eds. (Wiley, London, 2001), pp. 341-361. 3. [↵][9]1. C. S. Atwood 2. et al. , J. Biol. Chem. 273, 12817 (1998). 4. [↵][10]1. C. S. Atwood, 2. G. M. Bishop, 3. G. Perry, 4. M. A. Smith , J. Neurosci. Res. 70, 356 (2002). 5. [↵][11]1. R. Pluta 2. et al. , Neuroreport 10, 3615 (1999). # Cerebral Hemorrhage and Amyloid-β {#article-title-2} The view presented by Atwood and colleagues that vascular β-amyloid (Aβ) accumulation may play a protective role after vessel injury in the central nervous system carries with it the implication that any anti-Aβ therapy has the potential to compromise the blood-brain barrier. We believe, however, that the preponderance of current evidence would argue that cerebral amyloid angiopathy (CAA) is inherently pathological ([1][12], [2][13]) and that our Brevia article is best interpreted in this light. Atwood et al. suggest that local insults to the vasculature may lead to the rapid accumulation of sealant Aβ in the vessel wall. This is an intriguing hypothesis, but so far lacks experimental evidence. In fact, there are several independent observations that would argue against such a view. For example, App -null mice do not exhibit spontaneous cerebral hemorrhages, nor do they have a leaky blood-brain barrier ([3][14]). Furthermore, in our study, Aβ immunotherapy did not induce cerebral bleeding in mice that lack CAA, arguing against a role of submicroscopic Aβ deposits as seals. Finally, our recent studies suggest that vascular amyloid has a neuronal origin and is not locally produced in response to vessel wall injury. Rather, neuron-derived Aβ is transported extracellularly over substantial distances to the vasculature, where it is cleared by periarterial drainage or by transport through the blood-brain barrier into the blood ([9][15], [5][16]). Deposition of vascular amyloid has been shown to lead to a loss of smooth muscle cells ([2][13]), and we agree that its removal by immunization may have resulted in the mechanical compromise of these weakened vessels. However, in our Brevia, we did not see at the light microscopic level a significant reduction in CAA after immunization ([3][14]). Although the mechanisms leading to hemorrhage are not well understood, this is the first reported side effect of vaccination in mice. Our study suggests that caution is warranted before undertaking Aβ immunotherapy in patients with substantial CAA. Further analysis of mouse models with CAA should help to determine the cause of CAA-associated hemorrhage as well as, hopefully, how to avoid it. 1. [↵][17]1. S. Greenberg , Neurology 51, 690 (1998). 2. [↵][18]1. D. Winkler 2. et al. , J. Neurosci. 21, 1619 (2001). 3. [↵][19]1. M. Pfeifer 2. et al. , unpublished data. 4. [↵][20]1. M. Calhoun 2. et al. , Proc. Natl. Acad. Sci. U.S.A. 94, 13287 (1997). 5. [↵][21]1. M. Meyer-Luehmann 2. et al. , Nature Neurosci. , in press. [1]: /lookup/doi/10.1126/science.1078259 [2]: #ref-1 [3]: #ref-2 [4]: #ref-3 [5]: #ref-4 [6]: #ref-5 [7]: #xref-ref-1-1 View reference 1 in text [8]: #xref-ref-2-1 View reference 2 in text [9]: #xref-ref-3-1 View reference 3 in text [10]: #xref-ref-4-1 View reference 4 in text [11]: #xref-ref-5-1 View reference 5 in text [12]: #ref-6 [13]: #ref-7 [14]: #ref-8 [15]: #ref-9 [16]: #ref-10 [17]: #xref-ref-6-1 View reference 1 in text [18]: #xref-ref-7-1 View reference 2 in text [19]: #xref-ref-8-1 View reference 3 in text [20]: #xref-ref-9-1 View reference 4 in text [21]: #xref-ref-10-1 View reference 5 in text